1. Field of the Invention
This invention relates to methods for screening drug compounds that bind to GABA.sub.a receptors. More specifically, it relates to the use of imidazoquinoxalines and imidazopyrimidines in such drug screening methods. It further relates to novel GABAa receptor subtypes which bind certain imidazoquinoxalines and imidazopyrimidines.
2. Description of the Related Art
.gamma.-Aminobutyric acid (GABA) is regarded as one of the major inhibitory amino acid transmitters in the mammalian brain. Forty years have elapsed since its presence in the brain was demonstrated (Roberts & Frankel, J. Biol. Chem 187: 55-63, 1950; Udenfriend, J. Biol. Chem. 187: 65-69, 1950). Since that time, an enormous amount of effort has been devoted to implicating GABA in the etiology of seizure disorders, sleep, anxiety and cognition (Tallman and Gallager, Ann. Rev. Neuroscience 8: 21-44, 1985). Widely, although unequally, distributed through the mammalian brain, GABA is said to be a transmitter at approximately 30% of the synapses in the brain. In most regions of the brain, GABA is associated with local inhibitory neurons and only in two regions is GABA associated with longer projections. GABA mediates many of its actions through a complex of proteins localized both on cell bodies and nerve endings; these are called GABAa receptors. Postsynaptic responses to GABA are mediated through alterations in chloride conductance that generally, although not invariably, lead to hyperpolarization of the cell. Recent investigations have indicated that the complex of proteins associated with postsynaptic GABA responses is a major site of action for a number of structurally unrelated compounds capable of modifying postsynaptic responses to GABA. Depending on the mode of interaction, these compounds are capable of producing a spectrum of activities (either sedative, anxiolytic, and anticonvulsant, or wakefulness, seizures, and anxiety).
1,4-Benzodiazepines continue to be among the most widely used drugs in the world. Principal among the benzodiazepines marketed are chlordiazepoxide, diazepam, flurazepam, and triazolam. These compounds are widely used as anxiolytics, sedative-hypnotics, muscle relaxants, and anticonvulsants. A number of these compounds are extremely potent drugs; such potency indicates a site of action with a high affinity and specificity for individual receptors. Early electrophysiological studies indicated that a major action of benzodiazepines was enhancement of GABAergic inhibition. The benzodiazepines were capable of enhancing presynaptic inhibition of a monosynaptic ventral root reflex, a GABA-mediated event (Schmidt et al., 1967, Arch. Exp. Path. Pharmakol. 258: 69-82). All subsequent electrophysiological studies (reviewed in Tallman et al. 1980, Science 207:274-81, Haefley et al., 1981, Handb. Exptl. Pharmacol. 33: 95-102) have generally confirmed this finding, and by the mid-1970s, there was a general consensus among electrophysiologists that the benzodiazepines could enhance the actions of GABA.
With the discovery of the "receptor" for the benzodiazepines and the subsequent definition of the nature of the interaction between GABA and the benzodiazepines, it appears that the behaviorally important interactions of the benzodiazepines with different neurotransmitter systems are due in a large part to the enhanced ability of GABA itself to modify these systems. Each modified system, in turn, may be associated with the expression of a behavior.
Studies on the mechanistic nature of these interactions depended on the demonstration of a high-affinity benzodiazepine binding site (receptor). Such a receptor is present in the CNS of all vertebrates phylogenetically newer than the boney fishes (Squires & Braestrup 1977, Nature 166: 732-34, Mohler & Okada, 1977, Science 198: 854-51, Mohler & Okada, 1977, Br. J. Psychiatry 133: 261-68). By using tritiated diazepam, and a variety of other compounds, it has been demonstrated that these benzodiazepine binding sites fulfill many of the criteria of pharmacological receptors; binding to these sites in vitro is rapid, reversible, stereospecific, and saturable. More importantly, highly significant correlations have been shown between the ability of benzodiazepines to displace diazepam from its binding site and activity in a number of animal behavioral tests predictive of benzodiazepine potency (Braestrup & Squires 1978, Br. J. Psychiatry 133: 249-60, Mohler & Okada, 1977, Science 198: 854-51, Mohler & Okada, 1977, Br. J. Psychiatry 133: 261-68). The average therapeutic doses of these drugs in man also correlate with potency at the GABAa family of receptors (Tallman et al. 1980, Science 207: 274-281.). Thus, a radioreceptor assay, using tritiated diazepam, can serve as a tool for the discovery of novel therapeutics at the GABAa receptor.
In 1978, it became clear that GABA and related analogs could interact at the low affinity (1 .mu.M) GABA binding site to enhance the binding of benzodiazepines to the clonazepan-sensitive site (Tallman et al. 1978, Nature, 274: 383-85). This enhancement was caused by an increase in the affinity of the benzodiazepine binding site due to occupancy of the GABA site. The data were interpreted to mean that both GABA and benzodiazepine sites were allosterically linked in the membrane as part of a complex of proteins. For a number of GABA analogs, the ability to enhance diazepam binding by 50% of maximum and the ability to inhibit the binding of GABA to brain membranes by 50% could be directly correlated. Enhancement of benzodiazepine binding by GABA agonists is blocked by the GABA receptor antagonist (+) bicuculline; the stereoisomer (-) bicuculline is much less active (Tallman et al., 1978, Nature, 274: 383-85).
Soon after the discovery of high affinity binding sites for the benzodiazepines, it was discovered that a triazolopyridazine could interact with benzodiazepine receptors in a number of regions of the brain in a manner consistent with receptor heterogeneity or negative cooperativity. In these studies, Hill coefficients significantly less than one were observed in a number of brain regions, including cortex, hippocampus, and striatum. In cerebellum, triazolopyridazine interacted with benzodiazepine sites with a Hill coefficient of 1 (Squires et al., 1979, Pharma. Biochem. Behav. 10: 825-30, Klepner et al. 1979, Pharmacol. Biochem. Behav. 11: 457-62). Thus, multiple benzodiazepine receptors were predicted in the cortex, hippocampus, striatum, but not in the cerebellum.
Based on these studies, extensive receptor autoradiographic localization studies were carried out at a light microscopic level. Although receptor heterogeneity has been demonstrated (Young & Kuhar 1980, J. Pharmacol. Exp. Ther. 212: 337-46, Young et al. 1981, J. Pharmacol Exp. Ther 216: 425-430, Niehoff et al. 1982, J. Pharmacol. Exp. Ther. 221: 670-75), no simple correlation between localization of receptor subtypes and the behaviors associated with the region has emerged from the early studies. In addition, in the cerebellum, where one receptor was predicted from binding studies, autoradiography revealed heterogeneity of receptors (Niehoff et al., 1982, J. Pharmacol. Exp. Ther. 221: 670-75).
A physical basis for the differences in drug specificity for the two apparent subtypes of benzodiazepine sites was demonstrated by Sieghart & Karobath, 1980, Nature 286: 285-87. Using gel electrophoresis in the presence of sodium dodecyl sulfate, the presence of several molecular weight receptors for the benzodiazepines has been reported. The receptors were identified by the covalent incorporation of radioactive flunitrazepam, a benzodiazepine which can covalently label all receptor types. The major labeled bands have molecular weights of 50,000 to 53,000, 55,000, and 57,000 and the triazolopyridazines inhibit labeling of the slightly higher molecular weight forms (53,000, 55,000, 57,000) (Seighart et al. 1983, Eur. J. Pharmacol. 88: 291-99).
At that time, the possibility was raised that the multiple forms of the receptor represent "isoreceptors" or multiple allelic forms of the receptor (Tallman & Gallager 1985, Ann. Rev. Neurosci. 8, 21-44). Although common for enzymes, genetically distinct forms of receptors have not generally been described. As we begin to study receptors using specific radioactive probes and electrophoretic techniques, it is almost certain that isoreceptors will emerge as important in investigations of the etiology of psychiatric disorders in people.
The GABAa receptor subunits have been cloned from bovine and human cDNA libraries (Schofield et al. 1988, Nature 328,221-227; Garrett et al., BBRC 156, 1039-1045, 1989). A number of distinct cDNAs were identified as subunits of the GABAa receptor complex by cloning and expression. These are categorized into .varies., .beta., .gamma., .delta., .epsilon., and provide a molecular basis for the GABAa receptor heterogeneity and distinctive regional pharmacology (Shivers et al. 1990, Neuron 4, 919-928; Levitan et al. 1988, Nature 335, 76-79). The .gamma. subunit appears to enable drugs like benzodiazepines to modify the GABA responses (Pritchett et al. 1989, Nature 338, 582-585). The delta subunit is associated with GABAa receptors that do not respond to benzodiazepines but may respond to other related compounds (Shivers et al., 1989, op. cit.) The presence of low Hill coefficients in the binding of ligands to the GABAa receptor indicates unique profiles of subtype specific pharmacological action.
Drugs that interact at the GABAa receptor can possess a spectrum of pharmacological activities depending on their abilities to modify the actions of GABA. For example, the beta-carbolines were first isolated based upon their ability to inhibit competitively the binding of diazepam to its binding site (Nielsen et al., 1979, Life Sci. 25: 679-86). Although predictive about potency at a receptor, the receptor binding assay does not totally predict the precise biological activity of such compounds; agonists, partial agonists, inverse agonists, and antagonists can inhibit binding. When the beta-carboline structure was determined, it was possible to synthesize a number of analogs and test these compounds behaviorally. It was immediately realized that the beta-carbolines could antagonize the actions of diazepam behaviorally (Tenen & Hirsch, 1980, Nature 288: 609-10). In addition to this antagonism, betacarbolines possess intrinsic activity of their own opposite to that of the benzodiazepines; they become known as inverse agonists.
In addition, a number of other specific antagonists of the benzodiazepine receptor were developed based on their ability to inhibit the binding of benzodiazepines. The best studied of these compounds is an imidazodiazepine, (Hunkeler et al., 1981, Nature 290: 514-516). This compound is a high affinity competitive inhibitor of benzodiazepine and beta-carboline binding and is capable of blocking the pharmacological actions of both these classes of compounds. By itself, it possesses little intrinsic pharmacological activity in animals and humans (Hunkeler et al., 1981, Nature 290: 514-16; Darragh et al., 1983, Eur. J. Clin. Pharmacol. 14: 569-70). When a radiolabeled form of this compound was studied (Mohler & Richards, 1981, Nature 294; 763-65), it was demonstrated that this compound would interact with the same number of sites as the benzodiazepines and beta-carbolines, and that the interactions of these compounds were purely competitive. This compound was the ligand of choice for binding to GABAa receptors to discover older leads because it does not possess receptor subtype specificity and measures each state of the receptor.
The study of the interactions of a wide variety of compounds similar to the above has led to the categorizing of these compounds. Presently, those compounds possessing activity similar to the benzodiazepines are called agonists. Compounds possessing activity opposite to benzodiazepines are called inverse agonists, and the compounds blocking both types of activity have been termed antagonists. This categorization has been developed to emphasize the fact that a wide variety of compounds can produce a spectrum of pharmacological effects, to indicate that compounds can interact at the same receptor to produce opposite effects, and to indicate that beta-carbolines and antagonists with intrinsic anxiogenic effects are not synonymous. A biochemical test for the pharmacological and behavioral properties of compounds that interact with the benzodiazepine receptor continues to emphasize the interaction with the GABAergic system. In contrast to the benzodiazepines, which show an increase in their affinity due to GABA (Tallman et al., 1978, Nature 274: 383-85, Tallman et al., 1980, Science 207: 274-81), compounds with antagonist properties show little GABA shift (i.e., change in receptor affinity due to GABA) (Mohler & Richards 1981, Nature 294: 763-65), and the inverse agonists actually show a decrease in affinity due to GABA [(Braestrup & Nielson 1981, Nature 294: 472-474)]. Thus, the GABA shift in a receptor binding assay in addition to high affinity predicts generally the expected behavioral properties of the compounds.